Reply to Comment by H. Kitano and A. Maeda

Abstract

In our recent Letter, we reported that, in the metallic state of -(BEDT-TSF)2FeCl4 at TMI 1⁄4 8:3K T TFM 1⁄4 70K, the microwave conductivity 1 along the c axis is much more resistive than the dc conductivity dc, and a ferroelectric-like response is revealed in the dielectric constant 1, which is enhanced to the order of 10 . The unusual metallic state is identified by other experiments such as specific heat, NMR and ESR. From the absence of these anomalous features in the non-magnetic, isostructural GaCl4 salt, we pointed out the possible importance of the -d interaction for the ferroelectric-like anomaly as well as for the exchange-induced MI transition. In the preceding Comment, Kitano and Maeda (hereafter abbreviated as KM) insist that the probable metallic shift =n may be 4:25 10 5 obtained by eq. (1) (see below) applied to the depolarization peak, and hence 1 may result in a negative value. We point out here that the method to determine a metallic shift cannot be applicable to the depolarization peak in the present system, and the premise of KM, that the system might be in a usual highly-metallic state above TMI, is contrary to the experimental facts. First of all, we start to discuss on the anomalous behavior in =2f0 at TMI T TFM. We had carried out microwave measurements at 16.3GHz on about 20 single crystals before we submitted our paper. Every data has exhibited a saturation as shown in Fig. 1 of ref. 1 or a shallow minimum of =2f0 around 40–50K as depicted here in Fig. 1. Recently such minimum is also reproduced clearly in cavityperturbation measurements at 45GHz, which will be published elsewhere. According to the dc resistivity dc, which indicates a metallic reduction down to TMI, =2f0 could be expected to approach to zero with decreasing temperature, if the present system be a usual metal. In fact, in case of the GaCl4 salt, =2f0 decreases, in the same temperature range, in accordance with the dc resistivity as demonstrated in Fig. 2 of ref. 1. Thus, the enhancement of microwave loss is confirmed to be so intrinsic in the FeCl4 salt that the anomalous metallic state is realized above TMI. The anomalous microwave loss must be taken into account in a consistent way for the evaluation of the microwave skin depth mw using 1. As mentioned later, 1 becomes two-orders of magnitude lower than dc, so that mw is about 10 times larger than dc. Hence the system never enters a skin depth regime and still remains in a depolarization regime. From these results, it is concluded that the system responses quite anomalously to microwave in the highly metallic state as far as the dc conductivity concerned. This surprisingly important fact is little recognized in KM, which described shortly ‘‘due to some extrinsic origin’’. Secondly, we shall now look more carefully into the determination of the metallic shift from the observed depolarization peak. Whenever a depolarization peak in =2f0 is detected, one first tries to use a following relation to determine a metallic shift =n as KM also mentioned: